Shield Gets Closer to Full HA with RAP 3.1
October 21, 2008 Alex Woodie
With the launch of Receiver Apply Program (RAP) version 3.1 earlier this year, Shield Advanced Solutions has narrowed the gap between its remote journaling-based product and high availability (HA) products that have been established in the market for years. The capability to automate the switch-over process and reverse the flow of journaling is perhaps the most important new feature in RAP 3.1 bringing the product closer to feature parity with other HA solutions.
Shield launched RAP in April 2007 to provide an alternative to i OS high availability software, which the company felt was too feature rich and too expensive to adequately serve the needs of the majority of small and midsize AS/400, iSeries, System i, i5, and i-based Power Systems organizations.
RAP utilizes IBM‘s journaling and remote journaling technology to obtain changes made to database files, objects, and IFS files on production servers and apply those changes to a backup machine. The first version of RAP implemented the basics of a remote journaling-based disaster recovery system, and gave users an automated way of handling any changes that can be applied using the APYJRNCHG command. The second version, which shipped in December 2007, introduced new synchronization and auditing features that brought the product closer to established high availability products on the market.
With version 3.1, which Shield announced in mid-July, the company introduced more features commonly supported in other HA products, including an automated switch management process (misstated as “witch management” in Shield’s announcement, angering warlocks everywhere).
RAP’s automated switch management features will help customers perform all the things that need to be done before executing a role swap, turning the backup system into the production system, and vice versa. It’s very important for an HA product to automate these tasks–including cleaning up the journal receiver chains, renaming the local journal receiver on the target system, and other chores–because it would be very easy for an administrator to miss a crucial step during the pressure-packed moments immediately before or after a disaster. In fact, one could make the argument that without an automated switch process, such a product is just a DR product, and not true HA.
Other new features in RAP 3.1 include support for independent auxiliary storage pools (iASPs), which are increasingly being used among i OS shops. Shield says it has rewritten RAP to take advantage of the iASP technology, and that iASP support is particularly useful for segregating customer data in hosted environments.
RAP 3.1 also brings new notification capabilities. Administrators can now receive e-mail alerts of “certain activities within the replication process,” Shield says. This release also introduces a new status screen that shows administrators information about the journaling system, such as current apply and last applied.
Keeping production and backup systems in synch will hopefully be easier thanks to new auditing features in this release. Shield is now able to compare the journal identifier (JID) on the source (production) and target (backup) systems, and highlight any mismatches for the administrator, who is also given the option of viewing the related object on the target system.
Last but not least, RAP 3.1 introduces new tools for cleaning up journal receivers. By being able to specify how many days’ worth of journal receivers to keep online, administrators will be able to minimize the product’s impact on DASD.